Remove Object API

A production Remove Object API that erases unwanted people, logos, wires, stickers and clutter from a photo and reconstructs the background. One HTTP call. $0.012 per image (12 credits). 300 free credits, no credit card.

Quick start — Remove Object API

Sign up, copy your key, and POST the image. Describe the thing to erase. The endpoint returns a generation id; poll until status=completed, then download the cleaned photo. White in an optional mask is the region to reconstruct.

curl -X POST https://api.pixelapi.dev/v1/image/remove-object \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "image=@product.jpg" \
  -F "prompt=remove the person in the background"

# Poll
curl https://api.pixelapi.dev/v1/image/GENERATION_ID \
  -H "Authorization: Bearer YOUR_API_KEY"

import requests
r = requests.post(
    "https://api.pixelapi.dev/v1/image/remove-object",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    files={"image": open("product.jpg", "rb")},
    data={"prompt": "remove the person in the background"},
)
job = r.json()
# poll GET /v1/image/{job['id']} until status == "completed"

import fs from "fs";
const form = new FormData();
form.append("image", fs.createReadStream("product.jpg"));
form.append("prompt", "remove the person in the background");
const res = await fetch("https://api.pixelapi.dev/v1/image/remove-object", {
  method: "POST",
  headers: { Authorization: "Bearer " + process.env.PIXELAPI_KEY },
  body: form,
});
const job = await res.json();

$ch = curl_init("https://api.pixelapi.dev/v1/image/remove-object");
curl_setopt_array($ch, [
  CURLOPT_POST => true,
  CURLOPT_HTTPHEADER => ["Authorization: Bearer YOUR_API_KEY"],
  CURLOPT_POSTFIELDS => [
    "image" => new CURLFile("product.jpg"),
    "prompt" => "remove the person in the background",
  ],
  CURLOPT_RETURNTRANSFER => true,
]);
$job = json_decode(curl_exec($ch), true);

What the Remove Object API is for

Catalog teams, marketplaces, and photo apps need a reliable way to delete a photobomber, a stray cable, a price sticker, or a logo without opening a desktop editor. The Remove Object API is a production REST endpoint that inpaints the hole so the surrounding texture, lighting, and perspective continue. You send one image and a short instruction; you get back a cleaned file at the original resolution.

It is not a background-cutout tool and it is not a text-eraser. Background cutouts belong on the Background Removal API. Captions, watermarks, and date stamps belong on the Remove Text API. Mixing those jobs here wastes a round-trip and produces a weaker fill.

Pricing comparison

Live charge is $0.012 per image (12 credits). That is the number /v1/pricing/live returns today — not a list price that drifts from the meter.

New accounts receive 300 free credits on a 24-hour trial, no card. That is enough to run a real catalog sample before you pay. After the trial, credits are prepaid; failed jobs that never produce an output are not supposed to keep the charge — check the dashboard if a delivery looks empty.

Production notes

• Auth: Authorization: Bearer YOUR_API_KEY from the dashboard.
• Input: multipart file field image (required). Optional prompt and optional binary mask.
• Output: poll GET https://api.pixelapi.dev/v1/image/{id} until completed, then fetch output_url.
• Keep the subject you care about out of the mask. Over-masking eats product edges; under-masking leaves a halo of the object.
• For a logo printed as lettering on a flat wall, prefer Remove Text. For a 3-D sticker or a person, stay on this endpoint.

Use cases and related pages

Teams wire this into ingest so every SKU is clean before it hits the storefront. Real-estate pipelines drop trash bags and cars. Event photographers drop photobombers. Fashion lookbooks drop clamps and tape.

Browse industry pages: ecommerce, fashion, real estate, photography. How-tos: remove a photobomber, remove people, remove power lines, watermark vs object. Alternatives: Cleanup.pictures, Magic Eraser. Also best inpaint API and AI Image Editor API for instruction-based edits that change a scene rather than erase a region.

After cleanup, many pipelines call the Upscale API so the reconstructed pixels print cleanly, or the Background Removal API for a transparent PNG.

Frequently asked questions